Unmanned aerial vehicle three-dimensional path planning method based on multi-strategy improved grey wolf algorithm

Abstract

The invention discloses an unmanned aerial vehicle three-dimensional path planning method based on a multi-strategy improved grey wolf algorithm, and the method comprises the steps: initializing the parameters of the grey wolf algorithm, including the population scale, the initial number of iterations, the maximum number of iterations, a convergence factor, and an unmanned aerial vehicle constraint total cost function; using Sine chaotic mapping to initialize a grey wolf population; fitness values of individuals alpha, beta and delta of the grey wolf algorithm are calculated, and the positions of the grey wolf individuals are determined; updating the convergence factor a; updating the gray wolf position by adopting a Cauchy mutation operator; introducing a step length optimization algorithm based on a tangent algorithm to obtain the next iteration position of the grey wolf individual; and judging whether the maximum number of iterations is reached or not, if not, continuing iteration, and if yes, outputting a global optimal path. The improved grey wolf algorithm SC-GWO is adopted, the convergence speed is higher, the precision is higher, and the globally optimal solution is accurately locked during path optimization.

Description

Technical Field

[0001] This invention relates to a three-dimensional path planning method for unmanned aerial vehicles (UAVs), and more particularly to a three-dimensional path planning method for UAVs based on a multi-strategy improved gray wolf algorithm. Background Technology

[0002] As missions become increasingly complex, problems such as poor communication stability, low navigation and positioning accuracy, and high path planning complexity of UAVs have gradually emerged. Therefore, path planning has become a hot topic in UAV research. The goal of UAV path planning is to find an optimal flight path from the starting point to the target point, while achieving a balance between path length, energy consumption, safety, and time.

[0003] Path planning can be divided into classical algorithms and swarm intelligence optimization algorithms. Classical path planning algorithms include A* algorithm, Dijkstra's algorithm, and Artificial Potential Field algorithm. These algorithms have drawbacks such as high computational complexity and unreachable objectives in complex and high-dimensional environments.

[0004] Swarm intelligence algorithms possess strong global search capabilities and can be improved to achieve better results for different practical problems. Among them, the Grey Wolf Optimizer (GWO) algorithm simulates the social behavior and hunting strategies of grey wolves, featuring structural simplicity and low parameter dependence, and has been widely used in UAV path planning and obstacle avoidance in recent years. However, in UAV 3D path planning, due to the complexity of the flight space, the GWO algorithm suffers from problems such as poor population diversity, premature convergence, slow convergence speed, and susceptibility to local optima.

[0005] To address these issues, existing technologies have proposed a hybrid gray wolf optimization algorithm—CLGWO—which incorporates a perturbation factor and dynamic weights from the Lion Swarm Optimization (LSO) algorithm into the gray wolf position update process to avoid the algorithm getting trapped in local optima. Furthermore, improvements have been made to its trajectory evaluation function, convergence factor, and position update method, enhancing its application in UAV path planning. However, these solutions do not consider the smoothness of the planned path. Therefore, this invention proposes a UAV 3D path planning method based on a multi-strategy improved gray wolf algorithm. Summary of the Invention

[0006] Purpose of the invention: The purpose of this invention is to provide a UAV 3D path planning method based on a multi-strategy improved Grey Wolf algorithm, which reduces the slow convergence speed and tendency to get trapped in local optima in the later stages of the Grey Wolf optimization algorithm (GWO). The improved Grey Wolf optimization algorithm (SC-GWO) has a faster convergence speed and higher accuracy, and can accurately lock the global optimum during path optimization.

[0007] Technical solution: The present invention provides a UAV 3D path planning method based on a multi-strategy improved gray wolf algorithm, comprising:

[0008] Initialize the parameters of the Grey Wolf algorithm, including the population size n, the initial number of iterations t, the maximum number of iterations MaxIt, the convergence factor a, and the total cost function of the UAV constraint;

[0009] Initialize the gray wolf population X using Sine chaotic mapping. i ;

[0010] Calculate the fitness values ​​α, β, and δ of the individual gray wolf using the gray wolf algorithm to determine the position X of the individual gray wolf. α X β X δ ;

[0011] Update the convergence factor a;

[0012] The position of the gray wolf is updated using the Cauchy mutation operator;

[0013] A step size optimization algorithm based on the tangent algorithm is introduced to obtain the next iteration position of the gray wolf individual;

[0014] Determine if the maximum number of iterations has been reached. If not, continue iterating. If so, output the globally optimal path.

[0015] Furthermore, the Sine chaotic mapping is formulated as follows:

[0016] X n+1 =μ·sin(πX) n )

[0017] Where X n ∈[0,1],μ∈(0,1] are control parameters. When μ=1, it enters a completely chaotic state.X0 is the initial value, and n is the size of the gray wolf population.

[0018] Furthermore, the updated convergence factor 'a' includes introducing a sine function, as shown in the following formula:

[0019]

[0020] Where t is the current iteration number, and MaxIt is the maximum number of iterations for the population.

[0021] Furthermore, the Cauchy mutation operator updates the gray wolf's position using the following formula:

[0022] X new (t)=X(t)+η·Cauchy(0,1)

[0023]

[0024] Where X(t) is the global optimal solution in iteration t; X new (t) is the global optimal solution after the Cauchy mutation operator is introduced and updated in generation t; position η is the mutation weight; Cauchy(0,1) is the standard Cauchy random distribution at t=1; λ is the adjustment parameter; and MaxIt is the number of iterations.

[0025] Furthermore, the value range of the adjustment parameter λ is 30 to 100.

[0026] Furthermore, the step-size optimization algorithm based on the tangent algorithm obtains the next iteration position of the gray wolf individual, as shown in the formula:

[0027] X(t+1)=X(t)+Step*tanhθ

[0028] Where X(t+1) is the position of the gray wolf in the next iteration, step is the size of the movement step, and tanθ is used to control the direction of the algorithm.

[0029] Furthermore, the tanθ controls the direction of the algorithm. When θ is closer to π / 2, the tangent value is larger, and the obtained solution is farther away from the current solution, which is used for the algorithm to search in space and escape local optima. When θ is closer to 0, the tangent value is smaller, and the obtained solution is closer to the current solution, which is used for local exploration and to accelerate the convergence speed of the algorithm.

[0030] Furthermore, the formula for calculating the step size is as follows:

[0031]

[0032] Where e is the position of wolf α, f is the position of wolf ω, ef is the Euclidean norm of vectors or matrices e and f, rand is a randomly generated number between (0,1), and n is the population size.

[0033] A computer device includes one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the programs, when executed by the processors, implement the steps of the UAV 3D path planning method based on a multi-strategy improved gray wolf algorithm.

[0034] A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the UAV 3D path planning method based on a multi-strategy improved gray wolf algorithm.

[0035] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages:

[0036] (1) The present invention uses an initialization method based on Sine chaotic mapping to enhance the diversity of the population during the search process;

[0037] (2) Introducing a sine function to update the convergence factor improves the search performance of the algorithm. The search range should be as large as possible in the early stage to expand the search space and avoid the algorithm from getting stuck in local optima. The rapid reduction in the later stage improves the local exploitation capability of the algorithm, which is conducive to improving the convergence speed and accuracy.

[0038] (3) The position of the gray wolf is updated again using the Cauchy mutation operator. Gray wolves closer to the original value can improve the local development ability of the algorithm, while gray wolves farther away from the wolf pack can improve the global search ability of the algorithm. The improved gray wolf algorithm increases the small local search of the original position, reduces the probability of the algorithm getting trapped in local optima, and improves the accuracy of path planning. Attached Figure Description

[0039] Figure 1 This is a flowchart of the method described in this invention.

[0040] Figure 2 This is a graph showing the change of the convergence factor 'a' of the present invention with the number of iterations.

[0041] Figure 3 These are two 3D views for terrain three-dimensional path planning in embodiments of the present invention.

[0042] Figure 4 These are top views of two terrain three-dimensional path planning embodiments of the present invention.

[0043] Figure 5 These are two side views of three-dimensional terrain path planning in embodiments of the present invention.

[0044] Figure 6 These are iterative diagrams of two terrain cost functions in embodiments of the present invention. Detailed Implementation

[0045] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0046] like Figure 1 As shown, a UAV 3D path planning method based on a multi-strategy improved gray wolf algorithm includes:

[0047] (1) Initialize the parameters of the Grey Wolf algorithm, including the population size n, the initial number of iterations t, the maximum number of iterations MaxIt, the convergence factor a, and the total cost function of the UAV constraint;

[0048] (2) Initialize the gray wolf population X using Sine chaotic mapping. i where i = 1, 2, 3, ..., n;

[0049] The Sine chaotic mapping formula is as follows:

[0050] X n+1 =μ·sin(πX) n (1)

[0051] Where Xn∈[0,1],μ∈(0,1] are control parameters, and the system enters a completely chaotic state when μ=1; X0 is the initial value, set μ=0.97, X0=0.6, and n is the size of the gray wolf population.

[0052] (3) Calculate the fitness values ​​of individuals α, β, and δ in the gray wolf algorithm, and determine the position X of the gray wolf individual. α X β X δ ;

[0053] (4) Update the convergence factor a;

[0054] The convergence factor 'a' decreases linearly, and its value controls the speed and accuracy of the algorithm's convergence. To improve the algorithm's search performance, the search range should be as large as possible in the early stages to increase the probability of finding the global optimum, while the search range should be as small as possible in the later stages to accelerate convergence. Therefore, a sine function is introduced to better meet the computational requirements, as shown in the following formula:

[0055]

[0056] In the formula, t is the current iteration number, and MaxIt is the maximum number of iterations for the population.

[0057] (5) Update the position of the gray wolf using the Cauchy mutation operator;

[0058] X new (t)=X(t)+η·Canchv(0,1)

[0059]

[0060] Where X(t) is the global optimal solution in iteration t; X new(t) is the global optimal solution after the Cauchy mutation operator is introduced and updated in generation t; position η is the mutation weight; Cauchy(0,1) is the standard Cauchy random distribution at t=1; λ is the adjustment parameter, which takes values ​​in the range [30, 100], and MaxIt is the number of iterations.

[0061] After the wolf pack updates its position, it updates the position again according to the Cauchy mutation operator. Gray wolves closer to the original position can improve the algorithm's local exploitation ability. At the same time, gray wolves farther away from the pack can improve the algorithm's global search ability. This adds a small local search to the original position. This local search strategy reduces the probability of the algorithm getting trapped in local optima, while improving the algorithm's local exploitation ability.

[0062] (6) Introduce a step size optimization algorithm based on the tangent algorithm to obtain the next iteration position of the gray wolf individual;

[0063] The position of the Grey Wolf in the next iteration is:

[0064] X(t+1)=X(t)+Step*tanθ (4)

[0065] Where X(t+1) is the next iteration position of the gray wolf, step is the size of the movement step, and tanθ is used to control the direction of the algorithm. According to the mathematical properties of tanθ, the closer θ is to π / 2, the larger the tangent value, and the farther the obtained solution is from the current solution, which is beneficial for the algorithm to search the space and escape local optima; the closer θ is to 0, the smaller the tangent value, the closer the obtained solution is to the current solution, which is beneficial for local exploitation and accelerates the convergence speed of the algorithm.

[0066]

[0067] Where e is the position of wolf α, f is the position of wolf ω, ef is the Euclidean norm of vectors or matrices e and f, rand is a randomly generated number between (0,1), and n is the population size. This strategy overcomes the original GWO's inability to cope with getting trapped in local optima after position updates. Compared to the original GWO, by controlling the step size and direction of x(t) movement, the convergence accuracy of the algorithm in escaping local optima is improved.

[0068] (7) Determine whether the maximum number of iterations has been reached. If not, continue iterating. If so, output the globally optimal path.

[0069] Determine if t ≥ MaxIt. If t < MaxIt, let t = t + 1 and return (3) to continue iterating. If t ≥ MaxIt, the algorithm iteration ends and the global optimal value of the algorithm is output.

[0070] This invention validates the effectiveness of the improved Grey Wolf Optimizer (SC-GWO) algorithm through a UAV 3D path planning simulation experiment. The maximum number of iterations in all experiments was 200, and the dimension dim = 10. The experimental results were compared with those of the Grey Wolf Optimizer (GWO), Golden Jackal optimization (GJO), Sine Cosine Algorithm (SCA), and Sand Cat Swarm Optimization (SCSO) algorithms. This experiment included obstacles of varying numbers, heights, and locations. Specific information is shown in Table 1.

[0071] Table 1 Mountain Types

[0072]

[0073] Table 2 shows the results of different optimization algorithms after 40 runs under the two simulated terrains described above.

[0074] Table 2. Optimal, average, and worst values ​​of the total cost (cost) function for different algorithms.

[0075]

[0076] Table 2 shows that the SC-GWO algorithm in this paper reduces the optimal cost by 20.6%, 21%, and 15% respectively compared to the GWO, SCA, and SCSO algorithms; the average cost is reduced by 18.5%, 20.2%, and 16.4%; and the worst cost is reduced by 14%, 16%, and 15%. The results indicate that the SC-GWO algorithm has the optimal path and good adaptability in both mountainous environments.

[0077] like Figure 3 , Figure 4 and Figure 5 As shown, the path planning results of various algorithms in 3D view, top view, and side view are presented respectively. It can be seen that, except for the GJO algorithm, each algorithm can effectively avoid obstacles. The SC-GWO algorithm has the shortest path length, the highest path smoothness, and obtains the global optimum, indicating that the SC-GWO algorithm has higher efficiency and better performance when handling complex problems. Figure 6 As shown, the convergence of different algorithms was compared. It can be seen that the SC-GWO algorithm has the fastest convergence speed compared to the GWO, SCA and SCSO algorithms, and the convergence curve is relatively stable in the later stage, indicating that the SC-GWO algorithm has high convergence accuracy.

Claims

1. A method for three-dimensional path planning of a UAV based on a multi-strategy improved grey wolf algorithm, characterized in that, The method comprises the following steps: Initialize the grey wolf algorithm parameters, including population size n, initial iteration number t, maximum iteration number MaxIt, convergence factor a, and total UAV constraint cost function; Initializing grey wolf population X using Sine chaotic map i ; Calculate the fitness value of the gray wolf algorithm individual α, β, δ, determine the gray wolf individual position X α , X β , X δ ; Update the convergence factor a; Update the grey wolf position by using the Cauchy mutation operator; Introduce a step length optimization algorithm based on the tangent algorithm to obtain the next iteration position of the grey wolf individual; Determine whether the maximum iteration number is reached, if not, continue iteration, if yes, output the global optimal path.

2. The multi-strategy improved grey wolf algorithm-based unmanned aerial vehicle three-dimensional path planning method according to claim 1, characterized in that, The Sine chaos mapping has the following formula: X n+1 = μ · sin(πX n ) where X n ∈ [0, 1], μ ∈ (0, 1] is a control parameter, when μ = 1, it enters a fully chaotic state, X0 is the initial value, and n is the gray wolf population size.

3. The multi-strategy improved grey wolf algorithm-based unmanned aerial vehicle three-dimensional path planning method according to claim 1, characterized in that, The updated convergence factor a comprises introducing a sine function, and has the following formula: Wherein, t is the current iteration number, and MaxIt is the maximum iteration number of the population.

4. The multi-strategy improved grey wolf algorithm-based unmanned aerial vehicle three-dimensional path planning method according to claim 1, characterized in that, The Cauchy mutation operator updates the grey wolf position, and has the following formula: X new (t) = X(t) + η - Cauchy(0, 1) Where X(t) is the global optimal solution at t generation; X new (t) is the global optimal solution after updating at t generation after introducing Cauchy mutation operator; position η is mutation weight; Cauchy(0, 1) is the standard Cauchy random distribution at t = 1; λ is the adjustment parameter, and MaxIt is the iteration number.

5. The multi-strategy improved grey wolf algorithm-based unmanned aerial vehicle three-dimensional path planning method according to claim 4, characterized in that, The value range of the adjustment parameter λ is 30 to 100.

6. The multi-strategy improved grey wolf algorithm-based unmanned aerial vehicle three-dimensional path planning method according to claim 1, characterized in that, The step length optimization algorithm based on the tangent algorithm obtains the next iteration position of the grey wolf individual, and has the following formula: X(t+1)=X(t)+Step*tanθ Wherein, X(t+1) is the next iteration position of the grey wolf, step is the size of the moving step length, and tanθ is used to control the direction of the algorithm.

7. The multi-strategy improved grey wolf algorithm-based unmanned aerial vehicle three-dimensional path planning method according to claim 6, characterized in that, The tanθ controls the direction of the algorithm, when θ is closer to π / 2, the tangent value is larger, the obtained solution is farther away from the current solution, which is used for space search and jumping out of local optimum; when θ is closer to 0, the tangent value is smaller, the obtained solution is closer to the current solution, which is used for local development and accelerating the convergence speed of the algorithm.

8. The multi-strategy improved grey wolf algorithm-based unmanned aerial vehicle three-dimensional path planning method according to claim 1, characterized in that, The calculation formula of the moving step length size step is as follows: Wherein, e is the position of the alpha wolf, f is the position of the omega wolf, e-f is the Euclidean norm of the vector or matrix e and f, rand is a randomly generated number between 0 and 1, and n is the size of the population.

9. A computer device, comprising: The computer program is executed by the processor to realize the steps of the method for three-dimensional path planning of a UAV based on a multi-strategy improved grey wolf algorithm according to any one of claims 1-8.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to realize the steps of the method for three-dimensional path planning of a UAV based on a multi-strategy improved grey wolf algorithm according to any one of claims 1-8.

Images